What to buy after AI stocks plunge? Oppenheimer: Focus on the "return realization" line, top picks are Microsoft, Oracle, and SAP

According to Zhitong Finance APP, after last week's severe volatility in global tech stocks—especially the crash in AI computing power and semiconductor sectors—the AI investment frenzy appears to be shifting from pursuing capital expenditure and corporate budget expansion to screening large tech platforms capable of delivering significant returns on massive corporate AI investments.

An Oppenheimer survey of 449 IT purchasers, one of Wall Street’s financial giants, shows that AI agent-driven AI application software products are still vying for larger enterprise budgets, but clients are beginning to focus more on return on investment validation. The pace of expenditure expansion could slow in 2027, meaning the software industry as a whole may not see accelerated broad-based growth. As a result, the Oppenheimer analyst team has identified Microsoft, Oracle, and SAP as top-tier potential winners, corresponding to diversified monetization, investment in AI infrastructure and operational efficiency, and key AI applications embedded in core business processes, respectively.

After the dramatic sell-off and subsequent rebound in tech stocks, the central question regarding AI computing infrastructure and AI agent-driven AI application thematic investments has shifted from how much budget enterprises are willing to allocate, to which platforms can turn gains in client efficiency into sustained revenue and profits. Platforms that already possess a customer base, business data, and workflow entry points, and that can convert verifiable client benefits into new paid subscriptions, have a better chance to turn the AI adoption boom into sustained growth.

AI remains a priority area for corporate budgets, but as procurement decisions shift from experimental deployments to return-on-investment validation, the expansion pace of expenditures may slow in 2027, with growth gains likely to be more concentrated.

On the one hand, enterprises are willing to invest in AI infrastructure, but on the other hand, they prioritize tech spending that enhances operational efficiency and cuts costs. Thus, in Oppenheimer’s view, Microsoft (MSFT.US) boasts diversified monetization channels covering cloud services, office tools, and development tools; Oracle (ORCL.US) fits the demand for enterprises to expand AI infrastructure and increase efficiency through tech investments; and German-based software giant SAP (SAP.US) embeds AI into core business processes such as finance, procurement, and supply chain, making it easier for clients to verify ROI through efficiency improvements and cost savings. Last week, both Microsoft and SAP outperformed the Nasdaq 100 and Philadelphia Semiconductor Index, while Oracle remained under significant correction pressure.

From Tech Stock Stampede to Strong Rebound, Software Sector Is Seeing Divergence Within

On last Thursday’s “Tech Stocks Crash Day,” November 5, the Nasdaq 100 Index dropped 6.36%, marking the largest single-day decline since March 2020; the total market cap of component stocks evaporated by about $2.3 trillion, setting a single-day record for market cap loss. Over the same period, the Philadelphia Semiconductor Index plunged 12.25%.

After the Democratic Party won control of both houses of the U.S. Congress, investors worried about stricter AI regulation, data center approvals, and energy cost-sharing requirements. These concerns could delay project deliveries and postpone cash returns associated with trillion-dollar AI infrastructure tied to AI chips and large AI data centers. Policy concerns, coupled with high valuations and crowded positions, were amplified by trend-following and other systematic strategies cutting risk, resulting in record-breaking tech stock sell-offs.

On Monday, the market began revising its previously overly pessimistic expectations: the Nasdaq 100 rose 3.40%, the Philadelphia Semiconductor Index rebounded by 6.85%, and the VIX fell about 22.3% to 19.92. Morgan Stanley, UBS, and other institutions believe that some project delays do not necessarily signal the end of the AI capital expenditure cycle. However, Nvidia declined by about 4.2% against the trend, affected by company-specific investigation news.

As for software stocks closely related to AI applications, based on closing prices from October 30 to November 6, Microsoft rose 1.67% last week, SAP’s US ADR fell 0.67%—both clearly outperforming the Nasdaq 100’s about 4.91% decline and Philadelphia Semiconductor’s roughly 12.15% drop; Oracle dropped 11.15%, only slightly better than the Semiconductor Index but significantly underperformed the Nasdaq 100. Microsoft showcased contrarian gains; SAP showed relative resilience; by comparison, Oracle has yet to show equal defensive characteristics.

Moving from Trials to ROI Validation! Microsoft, Oracle, and SAP Viewed by Oppenheimer as Potential “AI Winners” in the Software Industry

On Monday, Oppenheimer cited a recent survey indicating that Microsoft (MSFT.US), Oracle (ORCL.US), and SAP (SAP.US) are enterprise software companies that could become “super winners in AI applications.”

This survey, which interviewed 449 long-term enterprise IT system purchasers, found that artificial intelligence is steadily increasing its share in budgets. However, whether incremental monetization can be realized depends on return on investment.

Analysts wrote: “We believe the survey results indicate that as enterprises shift from trial phases to validating ROI, the pace of AI spending expansion will slow in 2027. The respondents’ answers also challenge the market’s general expectation for a broad-based reacceleration in the software industry next year.”

“On the contrary, we believe that the beneficiaries of AI will be increasingly concentrated, while many companies in the industry will continue to face headwinds such as slowing growth, limited room for upward earnings revisions, and continued valuation pressures due to ongoing concerns about technological disruption. We continue to be bullish on AI as a long-term budget priority, as well as its multi-year monetization opportunities for a handful of leading platforms. However, investors may underestimate the time required to validate enterprise ROI, which means AI’s lift to growth could be slower—and more concentrated—than currently anticipated.”

Thus, the investment institution sees Microsoft, Oracle, and SAP as key winners. Microsoft is considered to have “diversified AI monetization paths and strong profit-margins protection”; Oracle is seen as being “best positioned to serve enterprise clients prioritizing AI infrastructure and efficiency-first investments”; SAP, with its “embedded AI strategy and integration within business workflows, is well positioned as clients prove spending justification through efficiency gains and cost savings.”

Other potential AI winners include Agilysys (AGYS.US), which has predictable growth, and Paycom Software (PAYC.US); by contrast, Oppenheimer highlights that the latter’s benefit comes from the ability of AI-driven efficiency improvements to foster the creation of new businesses.

AI Budgets Still Rising, But Truly Scarce Resource Is Verifiable, Sustained Positive Returns

Oppenheimer points out a key shift: enterprise procurement is moving from “testing AI” to “validating how much AI is worth.” The proportion of AI in budgets can increase even as spending growth slows: companies are still willing to invest, but broad deployment from pilots requires data integration, permission settings, reliability tests, and cost-benefit accounting.

Therefore, rapid model capability advances do not necessarily translate immediately into accelerated revenue across the software industry. On this basis, the firm judges that growth boosts in 2027 may be slower and more concentrated than market expectations, with some companies still facing decelerating growth, limited earnings upgrades, and valuation pressures due to technology substitution concerns.

In its public evaluation, Microsoft’s advantage lies in diversified AI monetization pathways and strong margin protection; Oracle matches the enterprise need for priority investment in AI infrastructure and efficiency improvement; SAP, by embedding AI into daily business processes, enables clients to prove procurement value through cost reduction and efficiency gains. Other potential beneficiaries include Agilysys (AGYS), with strong predictability in growth, and Paycom (PAYC), which may benefit from AI-driven efficiency and the creation of new enterprises. It’s noteworthy that this is a stock-picking logic emphasizing differentiated beneficiaries, not a uniform bullish view on all software companies.

Mechanistically, Microsoft can charge at multiple points, including computing, development tools, office applications, and agent management: Azure handles model run demand; GitHub and Microsoft 365 handle development and productivity tasks; both subscriptions and usage-based billing broaden revenue streams. In the latest quarter, Azure revenue grew by 48%, Microsoft 365 commercial cloud revenue increased by 17% as reported, and Microsoft’s cloud gross margin was 66%; the company also disclosed that the number of Copilot customers using usage-based billing credits grew by over 75% quarter-on-quarter.

According to Oppenheimer, these data underpin a multi-channel monetization strategy. Oracle covers both OCI computing power, enterprise databases, and cloud applications; last quarter, infrastructure cloud revenue rose 121% and cloud applications grew by 10%. The differing growth rates indicate that Oracle’s AI advantage still largely involves infrastructure expansion, not just application subscription growth.

SAP’s core strength lies in its access to enterprise business data and corresponding processes, permissions, and transaction rules. As AI agents begin handling procurement, invoicing, inventory, or financial tasks more efficiently, they need to access data, perform operations, and leave audit trails under authorization; SAP has embedded Joule into these processes, making it easier for efficiency improvements, error reductions, and operational cost savings to be realized. In Q3, the company’s cloud ERP suite revenue grew by 29%, with 45,000 cloud customers reportedly using its AI solutions.

This leads to a key growth narrative: all three companies share the competitive advantage of integrating model capabilities into clients’ existing workflows and then monetizing via subscriptions, usage fees, or platform expansion. Their investment value ultimately depends on whether client cost savings and revenue gains can support ongoing payments and supplier profit growth.